Is 0.55 too high for replace_rate given Reversi can have draw result? about reversi-alpha-zero HOT 3 CLOSED

I see. You just drop out the draw games when evaluating. So it doesn't matter how often the draw happens.

from reversi-alpha-zero.

As you said, it is difficult to decide a good replace_rate.

you just drop out the draw games when evaluating

Yes, but I don't know whether it is a good choice.
In my evaluation configuration, game number for evaluation is 200. That of the paper is 400.
Still a little false positives are likely to occur.

Just curious what rate of drawing you got.

• total evaluation: 928/25812 ≒ 3.6%
• first 1000 games: 48/1000 ≒ 4.8%
• latest 1000 games: 38/1000 ≒ 3.8%

Therefore, your 1/8≒12.5% is a little high rate.

from reversi-alpha-zero.

Yes, but I don't know whether it is a good choice.

I think it depends on whether "draw" is a often-happened result for two same level players. If it is less seen in real practice, then it is a good choice( however I would just ignore those draw games rather than counting them in the 200 number). And if not, then not. But That's basically the game's property, not training program's. I am not good at Reversi, so I can't tell.

your 1/8≒12.5% is a little high rate.

I see. I will wait and see my more result.

from reversi-alpha-zero.